the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Dec 2025
Natural disturbances increasingly affect Europe’s most mature and carbon-rich forests
Abstract. Europe's forests store nearly 40 PgC and provide a critical carbon sink of ~0.2 PgC yr-1, yet climate-driven disturbances increasingly threaten this capacity. Although disturbance rates from windthrow and bark beetle outbreaks have risen in recent decades, it remains unclear whether these events increasingly affect the oldest and largest trees, which store a disproportionate share of carbon. Here, we combine three decades of satellite-derived disturbance maps with spatially explicit data on forest age, biomass, and species composition to reveal patterns of structural selectivity across Europe. We show that natural disturbances have shifted toward older, carbon-rich stands, with disturbed forest area > 60 years old nearly tripling since 2010 (from 0.38 to 1.06 Mha). This structural shift is most pronounced in spruce-dominated regions of Central Europe (effect size = 1), where compound heat and drought events have amplified susceptibility to bark beetles. Biomass losses from natural disturbances in spruce forests increased eightfold between the early (2011–2016) and recent (2017–2023) periods. Trend-based projections indicate that, if current patterns of structural selectivity persist, natural disturbances could expose biomass carbon stocks equivalent to approximately 20 % of Europe’s contemporary forest carbon sink by 2040 (~0.05 PgC yr -1 or ~0.7 PgC cumulative). Our findings reveal a previously unquantified vulnerability: climate-driven disturbances increasingly affect forest structures with high per-hectare carbon stocks, amplifying disturbance-related carbon exposure and weakening the long-term effectiveness of Europe’s forest carbon sink. Adaptive management strategies that promote structural and compositional diversification in high-risk regions will be critical to stabilise forest carbon storage under continued climate change.
- Preprint
(1673 KB) - Metadata XML
-
Supplement
(1580 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-6288', Bogdan Brzeziecki, 18 Feb 2026
-
RC2: 'Comment on egusphere-2025-6288', Marina Rodes & Veronica Cruz-Alonso (co-review team), 17 Mar 2026
General comments
This manuscript presents a comprehensive large-scale analysis of forest disturbances across Europe, with a particular focus on how disturbance regimes interact with forest structural attributes such as age and biomass. By integrating multiple spatial datasets, including disturbance maps and forest structure information, the study provides a valuable and timely contribution to our understanding of how natural disturbances are affecting European forests in the context of ongoing environmental change. The continental-scale perspective, combined with the explicit focus on structurally mature and carbon-rich forests, makes this work particularly relevant for both the scientific community and forest management. Overall, the manuscript is clearly written and addresses an important and timely research question. However, several methodological and conceptual concerns arise regarding both the analytical approach and the actual scope of the study.
Firstly, although a wide range of natural disturbances affect European forests (e.g., fires, pests and diseases, windthrow, droughts), the authors focus exclusively on bark beetle outbreaks and windthrow. While it is understandable that some disturbances are difficult to detect reliably with satellite remote sensing and may therefore be excluded, such as droughts, the omission of forest fires is not justified in the manuscript. Fire and droughts are major disturbances in many parts of Europe, particularly in Mediterranean forests, and excluding it effectively removes Mediterranean forests from the analysis, as is evident in Figure 2a. Given that Mediterranean forests represent a significant proportion of Europe’s forest area, limiting the study to bark beetle infestations and windthrow narrows the geographical and ecological scope of the work. Under this view, the main messages of the manuscript appear somewhat overstated. Because the study excludes important disturbance types such as wildfires and drought-related mortality, and therefore omits large parts of Mediterranean forest systems, the conclusions may not be representative of European forests and natural disturbances as a whole. We encourage the authors to moderate the wording of the title, abstract, and discussion to better reflect the actual scope of the analysis and better justify why they decided to exclude some disturbance types.
Moreover, the manuscript clearly places a strong emphasis on bark beetle disturbances and on spruce dominated stands. For example, authors use a threshold to differentiate between age classes according to an ecological threshold based on bark beetles’ preference (L 141-142 “Given known ecological thresholds (e.g., bark beetles preferentially affecting spruce > ~60 years) […], we further grouped stands into broad age classes (1-60 years, > 60 years)”). Another example is how genus data are aggregated. In the manuscript genus data originally divided into eight classes are grouped in 3 groups: spruce, other needleleaf and broadleaf species. This grouping lacks ecological justification and effectively isolates spruce as the focal taxon while grouping all other genera into broad categories. As a result, the analysis appears to emphasize spruce‑specific patterns rather than providing a balanced comparison across taxa. For the reasons outlined above, we suggest either broadening the analysis so that it aligns with the current title of the manuscript or, alternatively, adjusting the wording to reflect the actual scope of the study. This could involve specifying that the work focuses primarily on Central Europe, explicitly referring to bark beetle and windthrow disturbances rather than using the broader term “natural disturbances”, and/or indicating that the analysis is particularly centered on spruce-dominated forests.
Secondly, to assess whether the age structure of disturbed forests changed over time, authors calculated the median 2010 age of disturbed pixels for the early (2011-2016) and recent (2017-2023) periods interpreting differences between periods as evidence that disturbances increasingly affect older (>60 years) or younger stands (1-60 years). While using a fixed point in time to calculate stand age is understandable, because forest age is fixed at its 2010 value, the discrepancy between the assigned age and the true age at the time of disturbance increases over time. This means that disturbances occurring in the 2017–2023 period are associated with systematically larger age errors than those occurring in 2011–2016. As a result, the two periods are not strictly comparable with respect to forest age, and disturbances affecting older forests in the later period may be systematically assigned to younger age classes. This temporal bias could be corrected by using a relative stand age to the beginning of each period but if not, it would need to be acknowledged and discussed. Also, the manuscript does not clarify how pixels that experience disturbances in both periods are treated. For example, if a stand older than 60 years is disturbed in 2011 and again in 2020, is it classified as >60 years in both cases, even if the 2011 event was a stand replacing disturbance that would reset its age? Additionally, it is unclear whether disturbance intensity is considered to distinguish stand replacing events from partial disturbances, which would have important implications for interpreting age-related patterns.
Similarly, the interpretation that disturbances increasingly affect high-biomass, carbon-rich forests may partly reflect changes in the underlying forest structure rather than changes in disturbance selectivity. Forest biomass across large parts of Europe has increased over recent decades, meaning that the availability of high-biomass stands in the landscape has also increased. If disturbances occur randomly with respect to biomass, an increasing fraction of disturbances would be expected to occur in high-biomass forests simply because such forests have become more widespread. The manuscript should clarify how this potential availability effect is accounted for when interpreting disturbance selectivity.
The manuscript refers to the ~100 km hexagonal units as “forest stands”. However, in forest ecology and silviculture the term “stand” typically refers to a relatively homogeneous forest unit at the scale of hectares to tens of hectares. The hexagonal units used here represent landscape-scale areas containing many different forest stands. Using the term “stand” for these units may therefore be misleading. We recommend using a term such as “landscape unit” to not create confusion. Also, we recommend reconducting messages like “natural disturbances increasingly occur in structurally homogeneous forests/stands” to “natural disturbances occur more frequently in landscapes where biomass distribution is more homogeneous”. Related to the forest landscape homogeneity, other methodological details should be clarified. For example, the analysis excludes hexagonal units with fewer than “50 valid disturbed pixels” (L168) to ensure stable estimates of disturbance selectivity. While this is understandable from a statistical perspective, it may also bias the analysis towards regions with relatively high disturbance activity. If the number of hexagons meeting this threshold differs between the two study periods, this could influence the comparability of the results for the two periods. Clarifying how many hexagons are retained in each period would help assess the robustness of the comparison. Second, the manuscript states that annual median CV values were calculated to assess continuous shifts in structural heterogeneity (L173). However, it is not entirely clear how the yearly calculation is done (see also related minor comments).
Finally, more detailed discussion of the study’s limitations, both in terms of data inputs and methodological choices, will be needed. For example, regarding disturbance data, the Forest Disturbance Atlas can miss gradual non–stand‑replacing disturbances, as acknowledged by its authors. These processes also shape forest dynamics, and their omission may affect the interpretation of disturbance–related patterns. For the stand age product, the original publication reports high uncertainty values, which could propagate into the analysis and should be addressed when discussing the robustness of the findings. Overall, the datasets selected are appropriate and represent the best options available for conducting this type of analysis, and the methodological framework is generally sound. Nonetheless, a more explicit discussion of the limitations associated with remote sensing products and analytical choices would help readers better assess the robustness and generality of the conclusions.
Specific comments
L 21, 147, 148, 397, 399:
The manuscript repeatedly uses terms such as “species‑specific” even though the analysis is based on maps where tree taxa are aggregated at the genus level that are grouped into broader functional groups. Because the study does not employ taxonomic information at the species level, this terminology is misleading and should be revised. The authors should adjust the wording to reflect the actual taxonomic resolution used (e.g., “genus‑level” or “taxon‑group–specific”).
L 59-60:
The statement that older, high-biomass stands may be particularly vulnerable to natural disturbances is currently supported by Jactel et al. (2017). However, that study primarily focuses on the role of tree species diversity in resistance to disturbances rather than on stand age or biomass per se.
L 70-71: “We examine whether disturbance impacts have shifted toward older, carbon-rich stands and how these patterns vary across dominant genera”
Consider removing this sentence since it is similar to what is expressed in L 74: “We quantify shifts in the age and biomass structure of disturbed stands, test”. In addition, the patterns analyzed are not across dominant genera but among Picea, rest of conifers and broadleaves.
L 102-104:
Disturbance information is aggregated to the 100 m grid using the fraction of disturbed pixels, while forest composition is represented by the most common genus within the cell. These two aggregation approaches differ substantially in how much within-cell information is retained. Using only the dominant genus may obscure compositional heterogeneity, particularly in mixed forests. The authors may wish to clarify the rationale for this choice or discuss its potential implications.
L 113-117:
Consider adding further description to the data sets not described above such as forest age, above ground biomass and forest fraction products. Including information about resolution, year(s) of the information provided, members and reference (in the case of forest fraction) is highly recommended.
L 119:
Disturbances previous to 2011 are not used, right?
L 122-123: “we selected all 100 m pixels with ≥30% forest fraction.”
Please, justify this threshold. Why ≥30% and not ≥50%?
L 122-123: “we retained only those where ≥50% of the forested area was disturbed...”
Please, justify this threshold and discuss possible implications.
L 148-150: “To investigate species-specific susceptibility to disturbance, we assessed the structural and cumulative biomass loss across three genus groups: Spruce, Other needleleaf (including Larix, Pinus, and other conifers), and Broadleaf (including Fagus, Quercus, and other broadleaf species).”
In L101-103 states that the original genus map (at 10 m resolution) was aggregated to 100 m using “a mode-based majority filter, assigning each 100 m cell the genus class that occurred most”. According to the original genus map there are eight genus classes. Later (L 148–150), the authors introduce a second reclassification step in which these eight classes are grouped into three broader categories (spruce, other needleleaf and broadleaf). It is unclear why this grouping was not applied prior to aggregation. Aggregating first using the eight-class scheme and only afterwards collapsing the classes can lead to inconsistencies. For example, a 100 m pixel composed of 30 % Quercus spp., 30 % Fagus spp., and 40 % Pinus spp. would be assigned to Pinus spp. under the majority filter. In the subsequent reclassification, this pixel would be labeled as “other needleleaf,” even though 60 % of the area corresponds to broadleaf genera. Please, justify or clarify the aggregation methodology.
L 151: “For each disturbed pixel, we used ensemble median aboveground biomass estimates”
Specify what “ensemble median above ground estimates” refers to. Besides, aboveground biomass from the ESA CCI biomass v6.0 provides biomass information for the years 2007, 2010, 2015, 2016, 2017, 2018, 2019, 2020, 2021 and 2022. Specify which years were used for each period, 2010,2015 and 2016 for 2011-2016 and 2017, 2018, 2019, 2020, 2021, 2022 for 2017-2023?
L 173: “To assess continuous shifts, we calculated the annual median CV for each ensemble member and aggregated these across the 20 biomass realisations”
Explain in more detail, please.
L249-251: “The contrast between stable harvest selectivity and shifting natural disturbance patterns suggests that climate-amplified biotic agents, rather than management changes, drive the observed structural selectivity.”
The comparison between harvesting and natural disturbances may be somewhat misleading because the two disturbance categories are not represented with the same completeness. While harvesting is quantified across all European forests, natural disturbances exclude important agents such as fire and drought-induced mortality. As a result, natural disturbances are likely underestimated in regions where these processes dominate, particularly in Mediterranean forests.
Technical corrections
L 36: “A large share of Europe’s forests originated from post-war planting campaigns and are entering maturity, during which carbon accumulation slows as stands approach saturation”
In this sentence, the use of “and” appears to be an error.
L49: “Living with bark beetles,” 2019; Weynants et al., 2024).”
Revise this citation, please.
L 51-53: “(“Korhonen K. T., Ahola A. et al. (2021) Forests of Finland 2014-2018 and their development 1921-2018,” n.d.; “Pulgarin Diaz J. A., Melin M. et al. (2024) Relationship between stand and landscape attributes and Ips typographus salvage loggings in Finland,”n.d.).”
Please, revise citation.
L 72: “(Besnad et al, n.d)”
Include year in this citation as it appears in references section.
L 72, 83, 516: refers to the citation Santoro and Cartus (2023) which is the v4 of the data set. However, according to lines 115 and 517 v6 is used.
L 82: “European Forest Disturbance Atlas v2.1.1 dataset”
Include acronym EFDA
L 100: “The map distinguishes eight classes: Larix, Picea, Pinus, Fagus, Quercus, other needleleaf,”
Please follow standard taxonomic nomenclature by using genus‑level notation (in italics) with “spp.” (e.g. Larix spp., Picea spp., Pinus spp., etc.)
L 113: “The annual disturbance fraction for each agent (harvest, natural disturbance)”
Specify that it is not natural disturbance but bark beetle or windthrow disturbance (there are other natural disturbances not included in the study)
L 114: (Besnard et al., 2021, n.d.)
Remove “, n.d.”
L 120: “2.4 Integration of the different Earth Observation data streams”
This section repeats the title used in section 2.3
L 134: “Where X and Y represent the 2010 age distributions of disturbed pixels in the early and recent periods, X ´ and Y´ are”
Correct the blank in “X ´”
L 149: “… three genus groups: Spruce, Other needleleaf (including Larix, Pinus, and other conifers), and Broadleaf (including …”
Revise capital letters in “other needleaf” and “broadleaf”
L 164-167: “We used harmonized disturbance and biomass datasets at 100 m resolution across Europe from 2011 to 2023, covering both natural disturbances (windthrow and bark beetle) and harvests. All data were aggregated to a 100 km hexagonal grid (EPSG:3035) to ensure consistency in regional comparisons. Pixels were included if forest cover exceeded 30%, and disturbance affected more than 50% of the forested area in a given year.”
This information is repeated. The same information is provided previously in lines 121-124 and later in lines 178-182.
L 614: Living with bark beetles: impacts, outlook and management options | European Forest Institute [WWW Document],2019. URL https://efi.int/publications-bank/living-bark-beetles-impacts-outlook-and-management-options (accessed 7.8.25).
Revise citation, please. Authors?
L 678: “Santoro, M., Cartus, O., 2023. ESA Biomass Climate Change Initiative (Biomass_cci): Global datasets of forest above-ground biomass for the years 2010, 2017, 2018, 2019 and 2020, v4. https://doi.org/10.5285/AF60720C1E404A9E9D2C145D2B2EAD4E”
Reference and link to version 4 while in the text authors refer to version 6. Please, correct according to the version used.
Marina Rodes-Blanco and Verónica Cruz-Alonso
Citation: https://doi.org/10.5194/egusphere-2025-6288-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 530 | 231 | 35 | 796 | 52 | 53 | 58 |
- HTML: 530
- PDF: 231
- XML: 35
- Total: 796
- Supplement: 52
- BibTeX: 53
- EndNote: 58
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Alba Viana-Soto
Henrik Hartmann
Marco Patacca
Viola H. A. Heinrich
Katja Kowalski
Maurizio Santoro
Wanda De Keersmaecker
Ruben Van De Kerchove
Martin Herold
Cornelius Senf
Europe’s forests store vast amounts of carbon, but climate-driven disturbances are becoming more frequent. By combining satellite records with information on forest age and structure, we show that recent disturbances increasingly affect the oldest and most carbon-rich forests, particularly spruce forests in Central Europe. This emerging pattern puts long-accumulated carbon at risk and may reduce the long-term climate benefits provided by Europe’s forests.
Europe’s forests store vast amounts of carbon, but climate-driven disturbances are becoming more...
General assessment
The subject of this work is the impact of natural disturbances (such as insect outbreaks and windstorms) and planned harvesting operations on the current condition and future development of European forests, in the context of their vital role in climate mitigation. The authors focused on basic tree stand’s features such as age, dominant tree species, aboveground biomass and spatial structure. In order to achieve the main paper’s goal, they first compiled and standardized relevant data from various sources and streams (in form of various thematic maps covering the entire area of Europe) and then processed them using advanced statistical methods and tools. They compared two periods: 2011-2016 (serving as a reference) and 2017-2023. Based on the obtained results, they concluded that, compared to the reference period, recently, not only the rate of disturbances and the damage they cause increased, but also the ecological profile of the affected stands changed significantly, i.e. current disturbances increasingly affect mature forest stands and forest areas distinguished by high carbon stocks (with particular emphasis on spruce-dominated forests). As pointed out by authors, structural shifts toward older, higher-biomass, and more homogeneous stands substantially amplify future carbon exposure and negatively affect the mitigation role of forests in relation to climate change. Their discussion highlights, among other things, the fact that reducing the future susceptibility of forests to increasingly frequent and intense disturbances "will likely require coordinated management of species composition, structural complexity, and spatial heterogeneity."
In general, the results obtained in this work are not surprising and confirm what has been known for a long time, such as the fact that older (and denser) and structurally less diverse stands are more susceptible to disturbances, like hurricane winds or harmful insects. Similarly, the idea of creating forest stands distinguished by diverse species composition and age structure is certainly not new and has been formulated many times before.
The above comments do not in any way lower my high assessment of this work, which, in my opinion, presents a very high scientific level. It is based on extensive empirical material. It employs a significant number of modern and advanced methodological approaches. The obtained results are presented in a clear manner, including aesthetically pleasing and easily readable figures. Thanks to the analyses conducted, the authors were able to quantitatively characterize the phenomena they studied, including estimating the magnitude of the reduction in the rate of carbon uptake by European forests by 2040, depending on the adopted scenario for future disturbance regimes.
Thus, in my opinion, the potential for significant improvement of this interesting, important and necessary work is not big. Possibly, a certain dissatisfaction may arise from the lack of an attempt to explain in more detail the reasons for the change in the characteristics of the stands affected by studied disturbance types that took place between the reference period and the present one. Was it the result of a change in the disturbance regimes and their vital parameters? Or the result of a change in approach from the forest management side, often forced by external circumstances? It would also have been possible to discuss in more detail the consequences of the obtained results from the point of view of what the optimal strategy for using forests should be for the purpose of mitigating climate change. I am thinking here, for example, of the recently strongly promoted concept of proforestation, i.e. withdrawing managed stands - including young and very young stands - from utilisation, placing them under strict protection and allowing them to “age” naturally. Does such a strategy make much sense, taking into account the results of this work and its forecasts for further developments?
Minor concerns
From a formal point of view, I wouldn't have too many critical comments about this paper, either. Perhaps in a few cases certain issues could be made a little clearer, especially from the point of view of a reader who is less familiar with the methods used in the work. Some examples from this field, including few editorial propositions, are given below:
Line 22. Add space after „area”.
Line 27. “expose” or maybe simply “release into the atmosphere”?
Line 38. “climate-sensitive” or maybe “climate-induced” or “climate-driven”?
Line 53-54. Maybe: “These trends are likely to persist as climate change continues” (to avoid repetition of “continue”).
Lines 55-56. Maybe: “such as windstorms Vivian and Wiebke (in 1990), Lothar and Martin (in 1999), and Klaus (in 2009).”?
Line 64. “are driven primarily by expanding spatial footprints”. Not very clear.
Line 72. What about: “with spatially explicit data on forest age…”?
Lines 85-87. It is not quite obvious for me how “the 30 m binary disturbance maps were aggregated to 100 m resolution”. Did you use information from nine 30 m x 30 m pixels (81,000 m2) per one 100 m x 100 m pixel (10,000 m2)?
Line 95. I am just not very sure if the term “disturbance” is appropriate in the case of (planned) harvesting activities. Maybe from the pure ecological point of view…
Line 95. “the forest age and biomass products” or just “the forest age and biomass data”?
Line 114. Is “20 members” clear enough?
Line 116. “forest fraction at 100 m pixel”?
Line 124. “Forest age was obtained from the 20-member GAMIv3.0 ensemble”. Repetition, see Line 114.
Line 128. indicated?
Line 134. “X ′ and Y′ are independent copies of X and Y”. Not very clear what is meant here by “independent copies”.
Line 143. “we summarised uncertainty across the 20-member ensembles”. I am not sure what is meant by “member ensembles”. Could you be more specific here?
Line 156. “ 20 biomass members”. Not very clear what does it mean.
Lines 164-167. Repetition. See lines 122-126.
Line 168. “each member of the 20-realisation biomass ensemble”. Not very clear, see comment to Line 156.
Lines 179-181. Repetition. See lines 122-126.
Lines 232-234. Partly repetition. Besides, this fragments fits better the Methods chapter. What does it actually mean: “higher values”? higher than what?
Line 348. Maybe “…the recent biomass loss scenario…”?
Fig. 5. “Biomass Loss: Natural Disturbance/Harvest (Late – Early)” or “Biomass Loss: Natural Disturbance/Harvest (Recent – Early)”?
Lines 428-430. This is certainly true, broadly speaking. However, there are studies, such as the long-term (started in 1936 and continued until today) study on permanent research plots in Białowieża National Park, NE Poland, that show that spruce is strongly declining also in stands characterized by significant age and species diversity. See Fig. 8 in Brzeziecki et al. 2020. Over 80 years without major disturbance .... J. Ecol. 108: 1138-1154. https://doi.org/10.1111/1365-2745.13367.
Line 443-444. “Together with the homogenising effects of salvage logging…”. Not very clear for me what do you mean by “homogenising effects of salvage logging”. For example, if salvage logging takes place in the initial period of insect outbreaks, then its effect is the creation of smaller or larger gaps in forest canopies, which lead to an increase in the structural diversity of the tree stands. These types of gaps make also possible introduction of more resistant tree species, either by natural regeneration or by planting.
Line 449. Tendency towards “accumulating high biomass over decades of growth” occurs also in the case of many natural forest types, for example in beech-dominated forest in Europe (see Schütz J.-Ph. 2002. Silvicultural tools to develop irregular…Forestry 75,4: 329-337.)
Line 500-501. “reducing susceptibility in high-risk stands”. It is not very clear what exactly is meant by that. Could you be more specific? What do you actually propose to do in the case of “high-risk stands”?